Detection of influential or outlier groups in multilevel modelling

In the context of multilevel modelling, I am trying to find methods for detecting if any of the groups strongly influence the results.

My particular dataset has multiple observations per participant and includes varying slopes and intercepts for participants.

I was hoping there was something out there like loo which could identify if removing any group (participant, in this case) from the dataset substantially impacts the posterior estimates.

I’ve seen some references to “leave one group out” but haven’t found any R functions that implement this simply.

You can do leave-one-group-out following the instructions for K-fold-CV in loo package vignette Holdout validation and K-fold cross-validation of Stan programs with the loo package • loo